In Zimbabwe, Sexual Relationships with Older Men Put Young Women at High Risk of HIV Infection
Among young people in Zimbabwe, the risk of HIV infection rises with the cumulative number of sexual partners and decreases with rising age at first intercourse.1 In addition, the risk of infection is six times as high among women as it is among men of the same age. Data from more than 4,400 men and women aged 15-24 indicate that the tendency of young women to have older partners--who are more likely to be infected--is a major factor in this differential.
Between July 1998 and January 2000, researchers surveyed households in rural eastern Zimbabwe to determine the risk behaviors associated with HIV infection among adults; all 15-54-year-old men and women were eligible to take part in the study. Literate participants were interviewed by structured questionnaire or by informal confidential voting;* nonliterate participants received face-to-face interviews. The interviews covered social and demographic information, sexual history (age at first sex and lifetime number of partners) and sexual behavior within the last two relationships in the past month (frequency of intercourse, type of relationships, frequency of condom use and partners' characteristics). All respondents provided blood for HIV testing.
Overall, 9,843 men and women participated in the study; 2,152 men and 2,276 women were younger than 25 and were included in these analyses. Male and female respondents did not differ in their median age at first sex (18.5 years), but men reported more lifetime partners than women: Among 24-year-olds, 51% of men said they had had five or more partners, compared with 5% of women. Young women reported having had sexual relationships with men who were a median of six years older, while young men reported having had sexual relationships with women who were a median of three years younger.
Fifty-one percent of young men and 39% of young women believed that their most recent partner had other partners. Young men were more likely than young women to have had casual or premarital sexual relationships (46% vs. 9% and 55% vs. 5%, respectively); however, men reported having sex less frequently in the previous two weeks than did women. Both men and women reported using condoms more consistently in their casual relationships than in relationships of 12 months or more (51% vs. 32% and 39% vs. 6%, respectively); condom use within marriage was 5%. Men were significantly more likely than women to use condoms consistently (odds ratio, 7.5). This may be partly explained by the fact that women were significantly more likely than men to be married (9.9), and that married respondents were less likely to report condom use with their most recent partners.
In bivariate analyses adjusted for age, the risk of HIV infection among 17-24-year-old women rose with the cumulative number of partners and number of years by which women were younger than their most recent partner, and declined with rising age at first intercourse. Moreover, women whose most recent partner had other partners and women who had consistently used condoms with their most recent partner had elevated risks of infection. For men, HIV risk rose with cumulative number of partners and age difference with the most recent partner.
In the first step of a reverse stepwise multivariate regression analysis, the researchers tested the effects of personal and behavioral factors on the risk of HIV infection among all respondents who had been sexually active in the previous year. Being female was the factor most strongly associated with HIV infection (odds ratio, 6.1). Among both women and men, the risk of infection rose with cumulative number of partners and declined with rising age at first intercourse.
When the analysis was limited to respondents who had had sex in the previous month and partner characteristics were entered into the analysis, risk rose slightly with each year by which the respondents were younger than their most recent partner (1.04 for both men and women). Age at first intercourse no longer had a significant effect on women's HIV risk, and no behavioral or personal characteristics continued to have an effect for men.
Further restriction of the analysis to respondents who had had sex in the previous two weeks did not change the effects of personal and behavioral factors; however, HIV risk was elevated among those who had used condoms consistently with their most recent partner (2.0 for both men and women) and rose slightly with each year by which the respondents were younger than their most recent partner (1.04)
The researchers conclude that the high prevalence of HIV among women in Zimbabwe occurs mainly because the cultural and social norms and economic realities of the region influence female adolescents to have sexual relationships with older men, who are more likely than adolescent males to be HIV-positive. They note that "although it is unrealistic to expect to alter the underlying socioeconomic context quickly, understanding its nature and influence on local patterns of sexual behavior should aid development of more relevant and, therefore, more effective HIV prevention strategies." The researchers suggest that public health programs that "highlight the extent of HIV infection among teenage girls," focus on getting men to stop having unprotected intercourse with commercial sex workers or "strengthen the socioeconomic position of young women" could help to reduce female teenagers' exposure to HIV.--J.Rosenberg
1. Gregson S et al., Sexual mixing patterns and sex-differentials in teenage exposure to HIV infection in rural Zimbabwe, Lancet, 2002, 359(9321):1896-1903.
*Participants were requested to record their responses to sexual experience questions on small preprinted cards, which they then placed in a locked voting box. This process is used to decrease participants' embarrassment and increase confidentiality.